基于滁州市TM数据的叶面积指数估算模型研建

以2010年TM影像为数据源,结合实测叶面积指数(LAI)数据,采用逐步回归方法,分析滁州市森林叶面积指数与植被指数关系并建立估测模型。结果表明:在0.01显著水平下,地面LAI和NDVI、RVI、SAVI的相关性分别为0.899、0.868、0.853;以NDVI为自变量构建的指数函数关系模型与LAI相关系数最高,相关性达0.839,LAI预测精度达78.96%;以NDVI、RVI、SAV为自变量构建的多元线性回归模型与LAI相关性达0.917,LAI估测平均精度达83.36%,符合森林资源监测要求。研究结果为使用遥感数据进行滁州市大面积森林质量监测、森林分布变化提供依据和技术支持。

Establishment of Leaf Area Index Estimation Model Based on TM Data of Chuzhou City

Based on the TM images of 2010 as the data source, combining with the measured LAI data, using stepwise regression method, the relationship between forest leaf area index and vegetation index in Chuzhou City was analyzed, the estimation model was established.The re-sults showed that:under 0.01 significant level, the ground LAI and NDVI, RVI, SAVI correlation were 0.899, 0.868, 0.853;the correlation coefficient between the index function relation model constructed with NDVI as independent variables was highest, correlation was 0.839, LAI prediction accuracy reached 78.96%;the correlation between multivariate linear regression model constructed with NDVI, RVI and SAV as in-dependent variables and LAI was up to 0.917, the average accuracy of LAI prediction was 83.36%, which is conform to the requirements of forest resources administration.The results can provide basis and technical support for large-area forest quality monitoring, forest distribution change in Chuzhou City using remote sensing data.